Viewing XML buffer data in Excel

Thanks for your response. I did notice that icon thingy in the error report. I am running WinXP Pro, 32-bit, SP3, with both Net 3.5-SP1 and Net 4-Extended installed. My default PNG program is Macromedia Fireworks which is a graphics editor closely coupled with DreamWeaver, a webpage editor.

Thanks

OK FOlks,

Another version hot of the keyboard for you.

Lots of Changes in this version.
Name changed to “XML data Viewer”
Moved application to an MDI format.
Added Double click events to the index buttons to centre trigger on display.
Added voltage cursors.
Added the ability to re-save a timeframe of the original XML file between time cursors.
Added Up down Buttons to the rotary controls
Changed S to s for seconds
Made data import library more robust to differing XML versions

I need someone to test the matlab and gnuplot exportsd for me please as i dont have these applicaions

Lygra … your offer of a video manual sonds like a great idea. do you want to have a go with what we have or wait for the final beta cut.

New version 1.0.0.5 attached

Screen Shot below

MainScreen.gif

Cheers Pete
XMLDataViewer_V1_0_0_5.zip (150 KB)

I’ve uploaded a new version that doesn’t use a PNG-compressed icon. I’d be interested in seeing if that solves your problem or not.

Thanks, the v99 file loads properly. I will experiment with it this week-end. It may prove helpful if you repeat your file location link with each update announcement post.

BenF has the best method of new releases, a new post for that version release. It keeps everybody focused upon the current release and reduces release confusion. You could start a new release post and include the file link in each new post. Bainsbunch could do the same and attach the new version zip in each new post.

Here are a few comments on this version of the application.

  • The spectrum analyzer is pretty cool, but it needs grids and axis labels.
  • The waveform flickers when it’s redrawn. You should use double buffering.
  • The waveform axis labels (voltage and time) should change when the offsets change, but they don’t.
  • The intention of the MDI model is to have multiple documents (multiple XML files open at the same time), not just multiple windows.
  • There’s a reason that Windows has theming support. As much as possible, you should attempt to adhere to the Windows appearance preferences selected by the user instead of overriding them. In short, not everyone wants their windows to be blue and pink.
  • You should be using Subversion and the Google Code release system. Immediate advantages are that you’d have an actual website instead of scattered forum posts, you’d have backups and the ability to revert any local accidents, other developers would be able to see your source instead of just guessing at what you’re doing from the binaries, etc. There are many tutorials such as this one - http://www.shokhirev.com/nikolai/programs/SVN/svn.html - for beginner developers to learn how to use source control. I’d be interested in looking at the most current version of your source, and Subversion is the easiest way to do it.
  • Excel export doesn’t work. It says that the provider is not registered on the local machine. This might be caused by compiling for AnyCPU or x64 instead of x86, though I didn’t check your binary target.
  • Here’s what your Matlab code looks like when it’s exported:

attenuation = x1; fileNumber = 'S001'; firmware = 'V3.61'; sampleCount = 4098; sampleDiv = 5 ms; timeDiv = 5 ms; timeRange = 0.8196; triggerIndex = 1374; triggerKind = 'EdgeRising'; triggerLevel = 78 mV; triggerMode = 'Auto'; triggerSens = 16 mV;

Attenuation, sampleDiv, timeDiv, triggerLevel and triggerSens should be numbers in scientific notation without any units or other characters. Otherwise it won’t run. The same applies to the Gnuplot export.

Good luck with development, and keep up the good work.

  • The spectrum analyser is pretty cool, but it needs grids and axis labels.
    - The waveform flickers when it’s redrawn. You should use double buffering.
    - The waveform axis labels (voltage and time) should change when the offsets change, but they don’t.
    <I’m not sure i understand what you are asking for here>- The intention of the MDI model is to have multiple documents (multiple XML files open at the same time), not just multiple windows.
    <well i thought that a container form could contain things, like other forms. Where is it specified that i cannot use an MDI container in the way i have>
  • There’s a reason that Windows has theming support. As much as possible, you should attempt to adhere to the Windows appearance preferences selected by the user instead of overriding them. In short, not everyone wants their windows to be blue and pink.
    <Perhaps, but this is not the first or the last application that is not doing it>
  • You should be using Subversion and the Goggle Code release system. Immediate advantages are that you’d have an actual website instead of scattered forum posts, you’d have backups and the ability to revert any local accidents, other developers would be able to see your source instead of just guessing at what you’re doing from the binaries, etc.
    <once the application is ready for its first release then i will publish the source. As a seasoned programmer of 20 odd years i don’t actually need a beginners tutorial on source control but than you for the reference. You don’t need to investigate the binaries to see what has changed just look at the history list embedded into the application>
  • Excel export doesn’t work. It says that the provider is not registered on the local machine. This might be caused by compiling for AnyCPU or x64 instead of x86, though I didn’t check your binary target.
    <I will look into this, it sounds like a real bug, not just a niggle about presentation>

Cheers Pete.

Thanks for another update. For more info on how the Nano uses cursors, refer to this video: youtube.com/watch?v=8T6OpWTEGjo.

I was expecting to view both the Data Window and the Waveform Window at the same time. I thought that you might keep the v1004 window ratios and just switch out different data as tabs where v1004 held the XML profile data. Maybe this helps to communicate my idea. The cursor data and the waveform need to both be viewed at the same time for maximum effectiveness.

I am out of time until this weekend, at which time I can provide more feed-back.

Thanks

Doh … OK MDI goes and back to tabs it is then, may be a squeeze getting it all into the top tabs but i will have a go.

CHeers Pete.

@Bainesbunch
WOW, I am astonished. I come back after few days and I see hugely improved application! :smiley:
Now I just need to find if it can be improved anything more :laughing:
Thank You Bainesbunch!

Hello Folks,

Lots of progress to report, the UART decoding is now working and here is a screen shot to see it in action. I am still working on the I2C and SPI as will try and get another release to you before the end of the weekend. I have moved the design away from a pseudo MID interface and back to a tabbed format, which makes seeing all the salient information much easier.

Cheers Pete.

MainScreen.gif

This is looking very user friendly now. Can’t wait to try it out. I especially like the Fake It button, I do need that on occasion. :smiley:

LOL :laughing: yeah it is there to create some I2C and UART data so i can debug the decoders. but if anyone has any real UART traces to test with it might be good to play with 96008n1 if you have some please and what it is supposed to contain. I might leave the button there for all those PPL who, like me, don’t have a nano to play with.

My quad is on its way according to the ordering system and the postal service claim to have it in their possession …

I am really hoping that someone like Ben will pick it up and run with it to create an export format for it like he did for the nano. The trouble is that the memory is limited and so it will most likely have to be binary and not XML. Either way as long as the schema is decided it can be imported into my client.

Cheers Pete

Hello folks,

Firstly sorry for those waiting for an update this weekend but I have had some issues with the I2C decoding but as you can see from the screen shot this is now working. I am not going to release another version until i have the SPI decoding as well this will be midweek hopefully.

Cheers Pete.

MainScreen.gif

Hello Folks here is version 1.0.0.6

In this version we have UART and I2C decoding seen in the screen shots.
Automatic saving and loading of configurations on open and close of the application. You still have the ability to save configuration to another file if you want to for loading at another time.

Some screen tidying up and moving of controls about moved it back to a tabbed layout away from the MDI format.

I am looking for more feedback now on what you would like to see included in the beta release.

Version 1.0.0.6 attached along with a default configuration file. this should be placed in the same folder as the application.

Cheers Pete.

LAScreen.gif
XMLDataViewer_V1_0_0_6.zip (156 KB)

Have been using v104 and am quite impressed with your overall layout. Here are a few of my constructive suggestions.

You seem to have everything under control except for the mouse pointer data. If you could display the mouse pointer data only while the mouse is on or very near a waveform data point (displays the nearest waveform data point data only), then that would be fantastic. With the current mouse pointer operation, most of the mouse pointer data is meaningless. Those cursor lines add unnecessary confusion.

If you can do the above, then another suggestion would be that when the user clicks on the mouse pointer location (while mouse pointer waveform data is present) then that becomes the reference for other mouse pointer locations that also display waveform data (now display and leave the cursor lines at this reference location). You could display both the current mouse pointer data and the relative mouse pointer data (in parenthesis, also no cursor lines for current mouse location). If you click the mouse again while waveform data is present, then that becomes the new waveform reference location. If you double click while on a waveform location then the mouse reverts back to pointer only and the parenthesis and reference offset data disappear. This would replace the need for a cursor system, and is used by many CAD programs.

The other suggestion would be to have a button to center the trigger location on the screen. I do believe that you have everything else covered. Oh yes, the help button should also display the version number.

The application is looking very good. Sorry it took me so long to review. :slight_smile:

Thanks very much.

I agree, I’ll work this into the next version.

Another feature I’ll work on eventually. This would be used to show time and voltage differentials, as well as mouse-driven zoom.

I’ll work on adding this, as well as some sort of home button that zooms out as far as possible, and some sort of auto button that fits the voltage scale automatically and uses FFT data to adjust the timescale to show perhaps five cycles.

Thanks again. Currently I’m working on integrating FFTW to support a detailed spectral view. Your feedback is appreciated.

In the future I suggest that you post feedback as comments under the project wiki’s “todo” page, or as “Type-Enhancement” issues in the issue tracker. Would you like me to add you as a project contributor?

Yes, please, thanks.

Excellent. You’re now in the project’s member list. I’ll add anyone else on this forum interested in testing the software.

A nit, but I thought you’d want to know. The About box shows v1.0.0.7

Thanks. Great work.

Paul

Well spotted, it is because I have the compiler auto generating the version when I re-compile

I am waiting on someone to create an XML output for the quad before I continue work on this project, but it is nod dead just sleeping :slight_smile:

Cheers Pete.